Aligning incentives in ICOs and token economies
In the new frenzy for digital assets, many entrepreneurs are debating if they should raise money through a token offering or not.
The decision is not that simple.
I’ve obviously been thinking non-stop about digital assets and ICOs for the past months, like many others.
In order to have a successful ICO and, most importantly, a smooth and successful development of the project, protocol, token or dapp, the needs and incentives of many different stakeholders (man do I hate this word, if you have a good alternative please let me know) must be carefully balanced.
It seems that many projects are carelessly dismissing this topic and just gunning for the most cash possible. That might make the developers rich in the short term, but makes everyone poorer in the long run.
In my opinion an ICO or token offering only makes sense if it can align incentives between every actor and ultimately make an application more successful than what it could achieve with a normal fundraising structure.
First, let’s see who’s involved in an ICO?
- Founding team
- Future employees
- Early investors
- ICO investors
- Future token investors
- End users of the dapp and thus of the token
- Equity holders in the offline-entity
Second, what levers can we act on in an ICO / token design?
- Total token supply
- Token allocation
- Number of coin offerings
- Price in pegged currency (-> valuation)
- Minimum rise
- Distribution schemes
Third, what does everyone want?
This is probably the hardest question, because right now it seems everyone just wants to make a buttload of money.
But let’s try to figure out what are the things we could maximize:
- Initial amount raised
- Future fundraises
- Token price over time
- Liquidity of the token
- Usage of the platform
- Revenues / dividends collected by token owners
- Participation from developers and others that can help the protocol
Obviously we can’t maximize all of the above at the same time, but the way we structure token sales and designs can have tremendous impact on the ones we choose.
Unfortunately even the most thoughtful ICO schemes that we’re seeing today are not really thinking that far into the future and only really trying to optimize the distribution and “fairness” of the ICO, which is an important but small part, for it changes very quickly after the ICO.
Let’s try to quickly examine current ICOs schemes:
- One-time fundraise: so all tokens are distributed at the beginning, often pre-product
- Some tokens reserved for team
- Some tokens reserved for advisors / partners etc.
- Fixed price per token or fixed valuation
- Maybe a cap on the amounts of tokens sold (-> so on capital raised)
In my view, current token sale models are completely suboptimal as they don’t really maximize ANY important thing!
And specifically only play to the advantages of early ICO investors with some sort of disregard for all of the others.
Let’s see how it fares on all of the dimensions we can tweak:
- Total amount raised: if there’s a cap, then it obviously doesn’t optimize this. If no cap then you are limiting the amount you can raise by doing one single ICO when there isn’t anything to show for your product.
- Token price over time: the thinking here goes that by limiting the supply and distributing it all on day-one will make the token value go up in time. This makes some sense, as long as the demand for that token exists and people want to own it, obviously. For that condition to happen there needs to be vast interest in the protocol and the products built on top of it, which are maximized in other aspects and not by supply dynamics. If instead the goal is for price stability over time so that people can use the protocol without worrying of becoming poor, then this clearly fails.
- Liquidity of the token: giving the token on day one to random speculators on the internet doesn’t guarantee a liquidity in the market, and actually might hinder it quite a lot as whales will just hodl as the platform becomes more successful.
- Usage of the platform: as above, if the token is fairly priced compared to the usage requirements and is liquid, people can use the platform.
- Revenues / dividends collected by token owners: this aspect is most impacted by token and economy design rather than ICO design, but it’s of utmost importance if we’re to gauge the future price potential of the asset. It also impacts usage and distribution of the token (eg. who wants to hold it).
- Participation from developers and others that can help the protocol: this can be impacted in the ICO by deciding how much of the token supply to distribute in exchange of work instead of in exchange for “cash”. The problem here is that reserving a random percentage of the token supply is a very complex guessing game as it’s pretty much impossible to predict how much the token will be worth in the future and in relation how much you can incentivize developers.
So, what’s the optimal structure?
Well the easy and obvious answer here is that there isn’t an optimal one, but a few easy changes that could help teams align incentives for everyone include:
This optimizes a lot of the things we were analyzing, as it gives the project proponent’s control over price / supply / risk for much longer.
Holding multiple fundraisers optimizes the amount raised and controls the token price over time.
Unfortunately, this is much harder to do when everything is immutable on a blockchain. I have a post in draft with a proposal on a structure, that I’ll try to publish soon.
Deciding who can own the token (at least in the early days, as it’s still a liquid instrument) might seem an “elitist” move but could actually be the easiest way to prevent whales and speculators messing up an economy and ecosystem.
Built-in, disclosed token economics!
If a token has a clear correlation to the success of the platform, more so than just a “well there will be more of a need for it”, then people will have greater incentives to own it and the proponents can tweak which kind of owners they’d prefer (users, speculators, etc.)
To incentivize participation and liquidity, I do think that most tokens would benefit from some sort of demurrage built-in.
Some easy models that could address some of the above:
Probably a topic for other posts (one for each!)
- Pre-sales for value-add owners.
- Multiple token types.
- Early helicopter drops to users.
This is such a nascent issue and theme that I think we need everyone’s contribution. If you have some comments please respond or email me, or better publish your own thoughts!
A number of very interesting posts have come out since I’ve started writing this post (been in draft for a while..):
Vitalik Buterin’s take:
Note: I mention the names of various projects below only to compare and contrast their token sale mechanisms; this…vitalik.ca
Albert Wenger’s take:
To keep track of all the new projects being launched in the space, I’m launching a newsletter called Token Economy very soon.
If you want to be kept up to date, you can subscribe on TokenEconomy.co.